Much of the convergence debate has focused solely on output. Recent empiric
al evidence suggests that crucial inputs, such as technology and capital, m
ay exhibit markedly distinct convergence patterns. We examine the convergen
ce characteristics of a two-sector nonscale model of growth that features p
opulation growth and endogenous technology. The model replicates key econom
ic ratios and speeds of convergence with relative ease. Most important, how
ever, is that capital and technology differ strikingly in their convergence
paths and speeds. The nonconstancy of the convergence rates and the nonpro
portionality of the endogenous variables during transition suggests further
refinements for the empirical tests of convergence.